![]() |
MeLOn
|
struct containing all information regarding the structure of a feedforward neural network More...
#include <AnnProperties.h>
Public Attributes | |
int | numLayers |
int | inputSize |
std::vector< int > | layerSize |
std::vector< ACTIVATION_FUNCTION > | activationFunction |
std::vector< int > | biasConnect |
std::vector< int > | inputConnect |
std::vector< std::vector< int > > | layerConnect |
bool | scaledInput |
bool | normalizedOutput |
struct containing all information regarding the structure of a feedforward neural network
std::vector<ACTIVATION_FUNCTION> melon::AnnStructure::activationFunction |
Vector containing the type of activation function for each layer
std::vector<int> melon::AnnStructure::biasConnect |
Vector containing an indicator for each layer that shows wether this layer has a bias or not
std::vector<int> melon::AnnStructure::inputConnect |
Vector containing an indicator for each layer that shows wether the network input is incident to this layer or not
int melon::AnnStructure::inputSize |
Size of the input vector
std::vector<std::vector<int> > melon::AnnStructure::layerConnect |
2D-Vector Indication the connections of each layer
std::vector<int> melon::AnnStructure::layerSize |
Vector containing the size of each layer
bool melon::AnnStructure::normalizedOutput |
Flag indicating if output should be normalized
int melon::AnnStructure::numLayers |
Number of network layers
bool melon::AnnStructure::scaledInput |
Flag indicating if scaling is used for input